Pharmacophore generation, atom-based 3D-QSAR, molecular docking and molecular dynamics simulation studies on benzamide analogues as FtsZ inhibitors

Abstract

FtsZ is an appealing target for the design of antimicrobial agent that can be used to defeat the multidrug-resistant bacterial pathogens. Pharmacophore modelling, molecular docking and molecular dynamics (MD) simulation studies were performed on a series of three-substituted benzamide derivatives. In the present study a five-featured pharmacophore model with one hydrogen bond acceptors, one hydrogen bond donors, one hydrophobic and two aromatic rings was developed using 97 molecules having MIC values ranging from .07 to 957 μM. A statistically significant 3D-QSAR model was obtained using this pharmacophore hypothesis with a good correlation coefficient (R² = .8319), cross validated coefficient (Q² = .6213) and a high Fisher ratio (F = 103.9) with three component PLS factor. A good correlation between experimental and predicted activity of the training (R² = .83) and test set (R² = .67) molecules were displayed by ADHRR.1682 model. The generated model was further validated by enrichment studies using the decoy test and MAE-based criteria to measure the efficiency of the model. The docking studies of all selected inhibitors in the active site of FtsZ protein showed crucial hydrogen bond interactions with Val 207, Asn 263, Leu 209, Gly 205 and Asn-299 residues. The binding free energies of these inhibitors were calculated by the molecular mechanics/generalized born surface area VSGB 2.0 method. Finally, a 15 ns MD simulation was done to confirm the stability of the 4DXD–ligand complex. On a wider scope, the prospect of present work provides insight in designing molecules with better selective FtsZ inhibitory potential.

Keywords:

• GTPase
• pharmacophore hypotheses
• 3D-QSAR
• molecular docking
• dynamics simulation

Abbreviations:

• FtsZ = filamentous temperature sensitive protein Z
• MIC = minimum inhibitory concentration
• 3D-QSAR = three-dimensional quantitative structure–activity relationship
• PLS = partial least square
• RMSD = root-mean-square deviation
• RMSE = root-mean-square error
• SD = standard deviation
• R² = correlation coefficient
• Q² = correlation coefficient for test set
• Glide XP = glide extra precision
• MM-GBSA = molecular mechanics-generalized born surface area
• MD = molecular dynamics
• MAE = mean absolute error
• σAE = standard deviation of the absolute error
• NPE = number of positive errors
• NNE = number of negative errors
• AE = average error
• AAE = average absolute error
• MPE = mean positive error

Acknowledgements

We would like to thank the Head, Department of Pharmaceutical Chemistry, JSS College of Pharmacy, Ooty, India and Innovative Informatica Technologies, Hyderabad, India for supporting the use of the software Schrödinger. We also thank Rt. A.C. Dash (Emeritus), Department of chemistry, Utkal University for helping us in scripting this article. We also thank Kunal Roy for helping us in using the software tool ‘Xternal ValidationPlus’.